There is widespread agreement that consent to research is a crucial prerequisite to research. But there is less agreement about what processes and documentation are appropriate in varying cultural and social contexts. The challenge for researchers is to establish procedures that are both ethically sound and culturally sensitive, although there may be times when these two requirements appear to be in conflict. The authors suggest that one way of resolving such situations is through careful and sustained community involvement in research.
During the last century, a number of medical research projects have been conducted without the knowledge or consent of those on whom the research was carried out.  Following widespread condemnation of such practice, it is now generally accepted that participants should consent to take part in medical research. This need is, for example, recognised in international human rights instruments,  as well as documented both in international guidance on research ethics,  and, in many countries, national forms of guidance and/or regulation. 
The need to obtain consent to conduct research with human participants is based on the fundamental ethical principles of 'respect for persons' and 'respect for human dignity'. These principles require that we do not use people merely as a means to an end, or act against their wishes. Instead we must respect their autonomy (their capacity to consider options, make choices, and act without undue influence from others) and the decisions that they make. It follows that we should seek their consent to participate in medical research.
Consent is the process by which prospective participants indicate their willingness to take part in research and give permission for researchers to undertake acts on them that are necessary to conduct the research. For such consent to be considered valid, it is now generally accepted that the prospective participant must: i) be appropriately informed about the nature of the research in which s/he is to participate; ii) adequately understand this information; iii) voluntarily decide to participate in the research; and iv) explicitly consent to participation.
While there has been general agreement about the need to meet these four conditions, there has been growing controversy over how such conditions can best be met. This is especially true in areas where a number of different cultures are represented, or in developing countries where, for example, prospective participants' beliefs about healing and causes of ill health may differ from those of medical researchers. Much of the controversy derives from an increasing awareness of the need for cultural sensitivity when conducting research. But the meaning of 'cultural sensitivity' is itself highly contested, which has implications for the implementation of consent procedures. 
Furthermore some have argued that for consent to be ethically acceptable, researchers must go beyond the minimal requirements that ensure their legal indemnity.  In other words, consent should be seen as more than a formal legal agreement, but as the outcome of an ongoing process whereby a prospective participant decides whether it is in their best interests to take part in research.
Informed consent: guidance and regulation
The need for formal consent to participate in research arose in the context of abuses of participants — particularly during the Second World War — and was initially articulated in the Nuremberg Code.  Such consent is only ethically acceptable if the participant decides about participation on the basis of knowledge and understanding of what the research will involve; this has developed into the concept of 'informed consent'.
For consent to participate to be appropriately informed, participants need to receive sufficient relevant and accurate information to enable them to make a genuine choice about participation. Various forms of international and national guidance and regulation set out issues that prospective participants should be informed about in detail.
The Declaration of Helsinki — adopted by the World Medical Association in 1964 — sets out the ethical principles that should be observed in research on human participants. It has become the cornerstone on which much national and international guidance is now based. The 2000 revision  requires that each potential participant must be adequately informed about:
- the aims of the study and methods to be used;
- sources of funding and possible conflicts of interest;
- institutional affiliations of researchers;
- anticipated benefits and potential risks;
- any follow-up (such as ongoing monitoring of possible side-effects);
- discomfort the study may entail; and
- the right to decline to take part, or to withdraw at any time, without reprisals.
The second set of international guidance that covers the field of research ethics is published by the Council of International Organisations of Medical Sciences (CIOMS) in collaboration with the World Health Organisation (WHO). A revision of its 1993 guidance was published in 2002,  and goes further than the Declaration of Helsinki, setting out 26 areas of which participants must be informed.
Conditions for consent to be valid
In the following sections we consider each of the four requirements for valid consent, taking account of the conditions that are likely to exist when research is being carried out in a developing country. We do not address the specific issues that arise when research is proposed in particularly vulnerable populations, such as children and refugees.
1) Providing appropriate and comprehensive information
A number of issues may arise when informing participants in developing countries, and/or from differing cultures, about research. For example, information that may be considered necessary or desirable in formally educated urban populations may be of little relevance in less formally educated or rural populations, or vice versa. For example, in a preventative HIV vaccine trial some populations may want highly technical information about the nature of the vaccine, whereas other populations may prefer information about the social impact of participating in the trial, such as risk of discrimination by other members of the community.
There is also a need to balance the use of extensive and detailed consent forms (which reflect the desire of organisations to protect participants and/or limit legal liability for possible harm caused by research) against ensuring that communities with limited biomedical and/or legal knowledge understand information about research.
For example, one Haitian researcher suggests that consent forms are too lengthy and have become increasingly complicated over the past 20 years. He states that "they appear to be more concerned about legal implications for sponsor agencies than…with the welfare of the volunteers. We cannot read them to volunteers because the only time a volunteer had a document like this read to him was when he was in a court of law and had to sign some kind of papers. So this is changing the trust relationship that we have with our participants and, therefore, we have to explain it step by step." 
It should also be remembered that in some cultures it might not be customary to provide certain forms of information, such as describing uncertainty about the effectiveness of the treatment being tested, or information about possible alternative treatments. In Vietnam, for example, it has been suggested that "it is unacceptable for a physician to openly express uncertainty with regard to what is the best treatment".  The need for comprehensive information also raises complex procedural questions such as who should decide which information is appropriate, and who is best placed to provide it.
As important as the content of the information provided to prospective participants is the way in which that information is provided. If consent is to be valid, information must not only be accurate, but provided in a culturally appropriate and understandable manner. Information may include unfamiliar concepts that are difficult to meaningfully translate in some languages, such as randomisation (selecting participants for different arms of a trial on a random basis) and placebos ('treatments' that have no effect, which are compared to a potentially effective substance or method).
Researchers have adopted different approaches to these difficulties, ranging from abandoning randomisation in research when it is thought to be impossible to obtain valid informed consent, to continuing with research despite concerns that participants have not understood the implications of using placebos. In other cases researchers have developed means of informing participants about unfamiliar concepts by using common examples, such as agricultural practices. 
The most recent CIOMS/WHO guidance sets out a detailed list of the responsibilities of investigators (Guideline 6), which includes seeking consent only after ensuring that the prospective participant has an adequate understanding of relevant information, and has had sufficient opportunity to consider whether to participate. But there are also subtle issues, such as the need to behave in a socially acceptable manner ('social desirability'), which may complicate this process - for example, making it difficult to ask questions - especially where there are marked differences in the social status of researchers and participants. This has led to the suggestion that researchers/clinicians and patients/participants might be 'paired' according to similar cultural and social values. 
The second condition for consent to be valid is that prospective participants adequately understand essential aspects of the research. Providing comprehensive information is not, in itself, considered sufficient grounds for assuming that the person fully understands that information. Consent procedures, therefore, may require some form of test of understanding, most commonly short yes/no or multiple-choice tests of factual information.
However, procedures for assessing participants' understanding have also been the subject of controversy, particularly when research is conducted in less formally educated, culturally diverse or developing country contexts, where the use of formal tests of knowledge is less appropriate. Questions that have been raised include: What level of understanding should participants demonstrate in order to be considered adequately informed? Are current tests measures of short-term memory rather than real understanding? Does the fact that participants often fail to remember information provided raise questions about real understanding? Who is best placed to decide whether a person has adequate understanding? 
The third condition that must be met for consent to be valid is that the person must freely consent to participate in the research (i.e. must not be subjected to undue influence or intimidation), and must be free to withdraw from the research at any stage without suffering prejudice. This represents the implementation of the core ethical principle of respect for persons, including respecting participants' autonomy.
A number of aspects of the social and cultural context in which research is conducted may affect a person's ability to voluntarily consent to participate in research, or their subsequent decision to stay in or withdraw from the research. For example, where there is a significant difference in social status between researchers and prospective participants, the latter may be reluctant to ask questions, or even to decline from taking part in recommended research. This may often be encountered in developing countries, where great respect is given to doctors and other healthcare workers.
Economic factors might also affect prospective participants' abilities to freely consent to research. For example, access to benefits that are provided to those taking part in research may act as a significant incentive to participate, especially in poorer areas where there is less likelihood of alternative access to healthcare. One Thai research participant stated: "The study staff gives good advice and when this project is over I hope I can enrol in another study. For that matter, I hope there will be new studies for me to participate in all the time. If there would be no more studies, I don't know if I would have the strength to go on, as I would not know where to get drugs outside of clinical trials." 
4) Formal consent
The final condition for valid consent is that there should be evidence of explicit consent to participate, which is usually done though signing a consent form in the presence of a witness. While this requirement has been widely accepted, there has been intense debate about the appropriateness of 'first person' consent (i.e. requiring that the participant him/herself provide consent) in all cultural and social contexts.
In some cases, community consent or consultation with leader(s) of the community or family is considered necessary before individuals are asked to participate in research. One Nigerian researcher commented: "There are really two levels. One is community and the other is individual."  The need for community consultation or consent is explicitly recognised in some forms of national or international guidance, as a safeguard against the abuse of vulnerable populations, or as an expression of specific cultural approaches to decision-making.
Although such guidance also stresses the need for consent from each participant in research, in some communities sexual or marital partners, or senior family or community members, may be considered to have authority to provide consent on behalf of others. This is referred to as proxy consent (as opposed to first-person consent). Some argue, however, that such cultural practices are ethically unacceptable, because they conflict with the fundamental ethical principle of respect for persons, and they carry the risk that participants will be enrolled in research against their will. 
To strike a balance between the ethical principles of respect for persons and respect for cultures, some suggest that in certain cultures it may be necessary to have multiple levels or forms of consent. For example, the consent of traditional leaders or community elders might be sought in order to enter a community or household,  and the explicit consent of individuals sought for participation in research. Another suggestion is that prospective participants might voluntarily consider inviting family members or community members to be part of the consent process.
The importance of obtaining valid consent as an essential prerequisite for research is widely accepted and entrenched in national and international guidelines for medical research. However, determining appropriate consent procedures, especially in culturally diverse and poorer settings, raises important and controversial issues.
The challenge for researchers is to establish procedures that are both ethically sound and culturally sensitive, although there may be times when these two requirements appear to be in conflict. One way of resolving such situations is through careful and sustained community involvement in research. Incorporation of communities as research partners, through community advisory boards for example, may go a long way towards finding solutions to apparent conflicts. Such practices would give added credibility to the research project, as well as provide researchers with the local knowledge necessary to design the most appropriate procedures to implement their projects.
Graham Lindegger is head of the School of Psychology, University of Natal, South Africa. Susan Bull is coordinator of SciDev.Net's dossier on the ethics of medical research.
 Examples include the universally condemned research conducted during the Second World War by Nazi physicians and scientists, in response to which the Nuremberg Code was formulated. Another example is research conducted on the progress of syphilis in Tuskegee, United States, where African American participants were denied treatment and enrolled in research without their knowledge or consent; an apology was subsequently issued by the White House.
 For example, article 7 of the International Covenant of Civil and Political Rights (1996) United Nations.
 For example the Ethical Guidelines for Biomedical Research on Human Subjects (2000) Indian Council of Medical Research; further examples are in the 'key documents' section of the dossier.
 Bayer R (1994) AIDS Prevention and cultural sensitivity: are they compatible? 84(6), 895-898.
 Lindegger G & Richter L (2000) HIV vaccine trials: critical issues in informed consent. South African J Science 96, 313-318.
 In 1947 the Nuremberg Military Tribunal delivered its verdict in the case of United States v Karl Brandt et al, in which twenty-three scientists and physicians stood trial for crimes committed before and during the Second World War. Included within the decision of the Tribunal was a statement defining legitimate medical research which subsequently became known as the Nuremberg Code.
 International Ethical Guidelines for Biomedical Research Involving Human Subjects (2002) Council of International Organisations of Medical Sciences (CIOMS) in collaboration with the World Health Organisation (WHO).
 Pape JP. Testimony before National Bioethics Advisory Committee (NBAC), Herndon, Virginia, February 29, 2000. Meeting transcript 32, as cited in Ethical and policy issues in international research: clinical trials in developing countries. Volume I: report and recommendations, p49 (2001).
 Love RR and Fost NC (1997) Ethical and regulatory challenges in a randomised control trial of adjuvant treatment for breast cancer in Vietnam. Journal of Investigative Medicine 45(8) 423-31.
 Preziosi MP et al (1997) Practical experiences in obtaining informed consent for a vaccine trial in rural Africa. New England Journal of Medicine 336(5) 370-373.
 Veatch R (1995) Abandoning informed consent. Hastings Centre Report.
 Lindegger G & Richter L (2000) HIV vaccine trials: critical issues in informed consent. South African J Science 96, 313-318.
 The HIV Netherlands Australia Thailand Research Collaboration: a model for HIV-AIDS clinical research in a developing country. (2000) UNAIDS, Geneva.
 National Bioethics Advisory Commission (2001) Ethical and policy issues in international research: clinical trials in developing countries. Volume II: commissioned papers and staff analysis, pC-17.
 For example Nuffield Council on Bioethics (2000) The ethics of research related to healthcare in developing countries. Nuffield Council on Bioethics, London, UK.
 Rugemalila JB & Kilama WL (2001) Proceedings of the seminar on health research ethics in Africa. Acta Tropica 78, S1-S126.